H‐NS‐like paralogues encoded by self‐transmissible plasmids (Shintani et al. 2015) and bacteriophage (Skennerton et al. 2011) can downregulate the expression of CRISPR‐cas loci, allowing the mobile genetic element to evade the host immune system (Dillon et al. 2012; Lin et al. 2016; Medina‐Aparicio et al. 2011; Pul et al. 2010). The LysR‐like transcription factor LeuO overcomes H‐NS‐mediated repression of the CRISPR‐cas locus, but the leuO gene is itself silenced by H‐NS, and its paralogues (Dillon et al. 2012; Medina‐Aparicio et al. 2011; Pul et al. 2010). Stochastic upregulation of leuO transcription may provide a mechanism for overcoming silencing of the immunity function in some cells in the population that encounter plasmid or bacteriophage invaders.
1.45 H‐NSB/Hfp and H‐NS2: H‐NS Homologues of HGT Origin
Genes encoding proteins related to H‐NS are found in pathogenicity islands that have been acquired by HGT. Hfp/H‐NSB is an H‐NS‐like protein that is expressed by a gene in the serU pathogenicity island in the chromosome of uropathogenic E. coli (Müller et al. 2010; Williams and Free 2005). It can form heterodimers with H‐NS and may modulate its activity in helping UPEC to adapt to environmental conditions encountered during the infection process (Dorman 2010; Müller et al. 2010). The H‐NSB/Hfp protein is also encoded by a chromosomal island in E. carotovora but is absent from an island in enteropathogenic E. coli (EPEC) that is closely related to the serU island of UPEC (Williams and Free 2005). This association of a gene encoding an H‐NS‐like protein with a former mobile genetic element is reminiscent of similar associations with elements that are currently mobile such as self‐transmissible plasmids and bacteriophage (Section 1.44). The absence of the hnsB gene from the version of the serU island that is present in EPEC is intriguing, given that hnsB is both present and expressed in the corresponding island in UPEC. Perhaps the gene had performed its role once the EPEC island was established in the chromosome and it was subsequently lost in the absence of a selective pressure to keep it? If so, the different environmental circumstances experienced by EPEC and UPEC seem to have selected for retention of hnsB by the latter organism.
Enteroaggregative E. coli (EAEC) strains express, in addition to H‐NS and StpA, an H‐NS2 protein that is closely related to H‐NS. H‐NS2 behaves somewhat like H‐NS when the latter is in a complex with Hha: it targets A+T‐rich genes that have been acquired by HGT and silences them transcriptionally (Prieto et al. 2018). The amino acid sequence of H‐NS2 is similar to those of H‐NSB and Hfp, but differs from them in a number of respects. It does not exhibit the sensitivity to proteolytic turnover that is a characteristic of these H‐NS homologues and StpA (Prieto et al. 2018). It is possible, and plausible, that H‐NS2 and other ‘third homologues’ could form heteromeric complexes with H‐NS or StpA that have distinct activities from those of the homodimers. Certainly, H‐NS heterodimers with StpA have properties that are distinct from those of the homodimers (Johansson et al. 2001; Leonard et al. 2009), so expanding the number of interacting partners may represent a way of modulating NAP function (Beloin et al. ; Sonden and Uhlin 1996; Zhang et al. 1996).
1.46 A Truncated H‐NS‐Like Protein
The serU island in UPEC that encodes H‐NSB/Hfp also encodes a protein that resembles a truncated H‐NS. This is H‐NST and its gene is tightly linked to the hnsB gene in the chromosomal island. H‐NST consists of the first 80 amino acids of H‐NS and the corresponding island in EPEC encodes a closely related protein; EAEC also encodes a relative of H‐NST (Williams and Free 2005). H‐NST from UPEC can form a heterodimer with H‐NS and it can antagonise its activity as a transcription silencer. The corresponding protein from EPEC is much attenuated in its ability to interact with H‐NS and to attenuate its biological activity: a key substitution at residue 16 of the amino acid sequence seems to be responsible for this difference between the UPEC and EPEC H‐NSTs (Williams and Free 2005).
The action of H‐NST recalls that of the gene 5.5 protein that is encoded by bacteriophage T7. Like H‐NST, the gene 5.5 protein co‐purifies with H‐NS and antagonises the transcription silencing activity of H‐NS, presumably to the benefit of the phage (Liu and Richardson 1993). H‐NST and the gene 5.5 protein resemble one another in size and mode of action structure but not in amino acid sequence (Williams and Free 2005). The ability of H‐NST from EPEC to inhibit H‐NS activity has been exploited to explore the H‐NS− phenotype of Yersinia enterocolitica, a bacterium where H‐NS is essential (Baños et al. 2008). The essential nature of H‐NS in Y. enterocolitica probably reflects the absence of a paralogous protein such as StpA that can offset the severe phenotype associated with the loss of H‐NS. Expressing EPEC H‐NST ectopically in Y. enterocolitica titrates the transcription silencing of H‐NS, revealing that it has similar effects on global gene expression patterns to those seen in other Gram‐negative bacteria such as E. coli (Baños et al. 2008).
1.47 Hha‐like Proteins
H‐NST's close similarity to the oligomerisation domain of H‐NS and its ability to form heteromeric complexes with H‐NS is superficially similar to the relationship of H‐NS to the Hha family of proteins. Found only in the Enterobacteriaceae, these too mimic the oligomerisation domain of H‐NS in their structure but unlike H‐NST, their interaction does not interfere with H‐NS‐mediated transcription silencing. Instead it channels the negative influence of H‐NS to certain target promoters (Baños et al. 2009; Madrid et al. 2007).
Y. enterocolitica possesses just one housekeeping Hha‐like protein (called Hha), in contrast to other model organisms like E. coli and Salmonella that have both Hha and a closely related paralogue, YdgT. However, pathogenic strains of Yersinia express YmoA (Yersinia modulator) from a chromosomal locus. YmoA is a founding member of the Hha protein family that regulates virulence genes negatively in Yersinia spp. (Cornelis et al. 1991; de la Cruz et al. 1992). It does this by forming a complex with H‐NS in which H‐NS provides the DNA‐binding activity (Ellison and Miller 2006b). YmoA potentiates the transcription repression activity of H‐NS, targeting virulence gene promoters in Yersinia (Ellison and Miller 2006b). It shares this property with Hha itself and with the Hha paralogue YdgT (Nieto et al. 2002; Starke and Fuchs 2014). YdgT is a paralogue of Hha and shares with Hha an ability to form heteromeric complexes with H‐NS and StpA (Paytubi et al. 2004).
YmoA is structurally closely related to Hha, it mimics the oligomerisation domain of H‐NS (McFeeters et al. 2007), and it is turned over by Lon‐ and ClpXP‐mediated proteolysis (Jackson et al. 2004). YmoA and Hha each interact with an H‐NS dimer, stabilising the transcription‐silencing complex at target promoters (Cordeiro et al. 2015).
In the case of Salmonella, Hha‐like proteins target H‐NS to the major virulence genes in the SPI1 and SPI2 pathogenicity islands and on the Salmonella virulence plasmid, pSLT (Silphaduang et al. 2007; Vivero et al. 2008). They can also influence the DNA‐binding mode of H‐NS and whether this protein forms polymers along DNA or creates bridges between different segments of DNA (van der Valk et al. 2017). Hha and YdgT direct H‐NS towards horizontally acquired genes, causing them to be silenced preferentially (Aznar et al. 2013). In E. coli, Hha/YdgT also targets horizontally acquired genes via H‐NS/StpA binding, together with genes involved in the osmotic and carbon starvation stress responses (Ueda et al. 2013). Genes encoding Hha‐like proteins also occur on self‐transmissible